Vol. 57, No.1, January 1992
FERTILITY AND STERILITY
Printed on acid-free paper in U.S.A.
Copyright e 1992 The American Fertility Society
A controlled assessment of direct intraperitoneal insemination
Eduardo Campos-Liete, Ph.D.*t Mark lnsull, M.B., Ch.B. Stephen H. Kennedy, M.B., Ch.B.
J. Douglas Ellis, M.B., Ch.B. Ian Sargent, Ph.D. David H. Barlow, M.D.:j:
The Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospita~ Oxford, United Kingdom
Objective: To test the efficacy of direct intraperitoneal (IP) insemination and its effect on cellmediated immunity. Design: A prospective trial with each couple having one treatment (insemination) cycle and one control (timed intercourse) cycle performed in random order with the same ovulation stimulation in both cycles. Setting: Secondary and tertiary referral fertility clinics; university teaching hospital. Patients: Twenty-six infertile couples. Excluding pregnancy, only one couple did not complete the two cycles. Interventions: Ovulation induction in both cycles. Intraperitoneal insemination in the insemination cycles. Main Outcome Measures: These were pregnancy rate (treatment versus control) and mixed lymphocyte response (MLR) sensitivity (before and after direct IP insemination treatment). Results: There were four control and no treatment pregnancies. This was not a significant difference (odds ratio). Mixed lymphocyte responses in fertile subjects did not change during the menstrual cycle (Wilcoxon). There was no significant increase in MLR sensitivity to partner's cells after direct IP insemination treatment. Conclusions: This controlled study found no benefit from direct IP insemination in terms of pregnancies over control cycles. There was no evidence that direct IP insemination had increased Fertil Steril1992;57:168-73 cell-mediated immune response sensitivity to husband's cells.
Direct intraperitoneal (IP) insemination has been promoted as a form of assisted reproduction involving less technology than full-scale in vitro fertilization (IVF) or gamete intrafallopian transfer (GIFT). It was first reported by Forrler et al. 1 and has been applied particularly in cases of unexplained or cervical infertility. 2- 5 It has also been combined with intrauterine insemination (IUI).6-7
Received November 1, 1990; revised and accepted August 23, 1991. * Recipient of a Post-Graduate Education Federal Agency (CAPES) scholarship no. 6859, Ministry of Education, Brasilia, Brazil. t Supported by the Oxford Regional Health Authority Locally Organised Research Scheme, Oxford. *Reprint requests: David H. Barlow, M.D., Nuffield Department of Obstetrics and Gynaecology, Level 3 John Radcliffe Hospital, Headington, Oxford OX3 9DU, United Kingdom.
168
Campos-Liete et al.
Direct IP insemination
The procedure involves injecting motile spermatozoa, after washing and migration, into the pouch of Douglas (cul-de-sac) via the vagina in a woman who has undergone ovulation stimulation. Ovulation is induced by clomiphene citrate (CC) or gonadotropins, and the insemination is timed to coincide with ovulation. The method avoids exposure of the spermatozoa to the cervical mucus and the uterine secretions as well as facilitating their transportation into the peritoneal cavity, a site they are known to reach after sexual intercourse. 8 In bypassing the usual route of entry into the pelvis, it may be that these spermatozoa represent a greater antigentic stimulus that could lead to activation of female cellmediated immune defenses. When new infertility treatments are introduced, it is important that they should undergo validation in controlled trials. Where the treatment involves
Fertility and Sterility
subjecting the woman to ovulation induction and an invasive pelvic procedure, it is particularly important to demonstrate that the treatment is effective. This study assesses the efficacy of direct IP insemination in a controlled trial with each couple having a direct IP insemination cycle and a control cycle and examines the effect of direct IP insemination on the female cell-mediated immune sensitivity to partner's cells. MATERIALS AND METHODS Patients
Twenty-six couples attending the two infertility clinics at the John Radcliffe Hospital, Oxford, United Kingdom, participated in the trial. All had undergone full investigation of infertility, and all had normal findings on laparoscopy within the previous 2 years. Fourteen couples had unexplained infertility, 8 had cervical factor infertility, and 4 had male subfertility. Fifteen had primary infertility (57.6%) and 11 secondary infertility (42.3%). The mean age of the women was 34.3 (range 26 to 40) years. The duration of infertility ranged from 2 to 10 years, with 5 of the 26 couples (19.2%) trying to conceive for >5 years. Study Design
The study design was for each couple to have a treatment and a control cycle in random order, unless pregnancy occurred in the first cycle. A treatment cycle involved ovulation induction with monitoring and timed direct IP insemination. A control cycle involved the same ovulation induction with monitoring and timed intercourse. The ovulation induction was either CC (Serophene; Serono Laboratories, Welwyn Garden City, United Kingdom) 50 to 150 mg/d for 5 days from cycle day 2 or human menopausal gonadotropin (hMG, Pergonal; Serono Laboratories) two to seven ampules on alternate days from cycle day 7. The final duration of administration of hMG was determined by ultrasonic monitoring. Allocation of the women to the two forms of ovulation induction was determined solely by the clinic of origin of the woman. Seventeen women were from the clinic of J.D.E. and received hMG, and 9 women were from the clinic of D.H.B. and received CC. The monitoring and timing were managed by a single researcher (E.C.-L.) for all patients.
Vol. 57, No. 1, January 1992
Human chorionic gonadotropin (hCG, 5,000 IU) was administered when at least one follicle was ;:::17 mm diameter. Only when the decision to administer hCG had been reached was randomization to a treatment or control cycle determined by drawing a card. The couples were asked to refrain from sexual intercourse from the start of ultrasonic monitoring. In control cycles, couples were advised to have sexual intercourse 24 and 48 hours after hCG. In treatment cycles, direct IP insemination was performed 36 to 48 hours after hCG. In the treatment cycles, the peritoneal fluid (PF) environment was sampled 24 hours after direct IP insemination to attempt detection of the inseminated sample. The study design had been passed by the Central Oxford Research Ethics Committee. Semen Preparation
Semen samples were obtained by masturbation and were allowed to liquefy (up to 30 minutes). In 4 cases (15.4%), the sperm density was in the range 10 to 40 X 106/mL; in 14 (53.8%), it was 41 to 100 X 106/mL, and in 8 (30.8%) >100 X 106/mL. A minimum of 5% and a maximum of 65% progressive motility was observed. In 2 cases, the morphology was poor with only 15% aild 19% normal forms. In the other samples, there was 51% to 69% normal morphology by World Health Organization9 criteria. After counting, the semen sample was split and each washed twice in Ham's F-10 medium (GIBCO, Paisley, United Kingdom) including 10% heat-inactivated patient's serum. The pellet containing the sperm was overlayered with medium and incubated at 37°C for 30 minutes. The best of the swim-up samples was used for the insemination. The semen preparations were all carried out by a single researcher (E.C.-L.). Insemination
Under vaginal ultrasound guidance an 18-gauge needle was inserted into the pouch of Douglas (culde-sac) via the vaginal probe needle guide. The needle was targeted on a pool of PF and collection of a portion of the fluid attempted (maximum 5 mL). Where PF could not be aspirated, fluid was allowed to drip from the needle end, confirming that the needle tip was indeed correctly located in the peritoneal cavity. The spermatozoa were injected in 1 mL of medium under ultrasonic visualization. The patient was allowed to rest for at least 15 minutes after the insemination.
Campos-Liete et al.
Direct IP insemination
169
Sperm Recovery Procedure The needling procedure was repeated at 24 hours after the insemination to attempt retrieval of some inseminated sperm. Because we expected that the women would be reluctant to have extra nontherapeutic needling procedures, we limited these to direct IP insemination cycles only. Eleven women agreed to this additional procedure in their direct IP insemination cycles. The volume of recovered fluid was measured and the specimen separated into cellular and liquid components by centrifugation. After washing in phosphate-buffered saline (PBS), a few drops of the cellular suspension were examined under phase-contrast microscopy for the presence of motile spermatozoa and the rest of the suspension further washed in distilled water and spotted (3 spots of 6 J.LL) onto slides. The slides were air dried and fixed using absolute alcohol, mounted with glycerine jelly, and examined under phase-contrast microscopy to determine the number of spermatozoa in each drop. Mixed Lymphocyte Reaction Testing The female cell-mediated immune response was assessed using the mixed lymphocyte reaction (MLR) in which lymphocytes from the woman (recipient) and her partner or unrelated male (donor) are cultured together for 6 days. The donor (stimulator) cells are irradiated to prevent them from responding. The proliferation of recipient (responder) cells is assessed via the incorporation of tritiated thymidine into newly synthesized deoxyribonucleic acid, and in a graft response this would peak at 6 days. By comparing the MLR reactions between the woman and her partner and between the woman and an unrelated control enhancement or suppression of the woman's immune reactivity may be tested. 10 Mixed lymphocyte reaction responses were assessed before direct IP insemination, 1 day after insemination and 7 days after insemination. Couples undergoing IP insemination were assessed in CCstimulated cycles (n = 8) or hMG-stimulated cycles (n = 17). A control group of fertile couples were assessed during natural menstrual cycles (n = 6) to determine if the major hormonal transitions of the menstrual cycle affected MLR responses. Each of the control couples had proven fertility (a child in the past 3 years), and the women had regular ovulatory menstrual cycles of 26 to 32 days' duration. They were not using hormonal contraception. Both partners gave blood samples on the 1st
170
Campos-Liete et al.
Direct IP insemination
day of the menstrual cycle, in the late follicular phase, and in the midluteal phase. Technique of MLR The peripheral blood samples were diluted 1:1 with PBS (10 mL) containing 5 J.Lg/mL heparin. The mononuclear cells were separated on Ficoll-Paque (Pharmacia, Upsalla, Sweden) and resuspended at 1 X 106 cells/mL in RPMI 1640 medium (Flow, Irvine, United Kingdom) with 25 mM Hepes and Lglutamine (GIBCO), containing antibiotic (50 J.Lgf mL penicillin and 50 J.Lg/mL streptomycin) and 10% pooled heat-inactivated AB serum from healthy donors. Quintuplicate 200-J.LL cultures were set up in flatbottomed 96-well microtiter plates with cellular mixtures of responder and stimulator cells, ratio 1: 1. The stimulator cells from husband and unrelated control were inactivated by irradiation (2,000 rad). The cell cultures were incubated at 37°C and 5% C0 2 in air. Blastogenesis was evaluated 5 days later, by adding 0.5 mCi [3H]thymidine to the wells. After a further 24-hour incubation, the cultures were harvested onto glass fiber filters that were allowed to dry. The filters were transferred to 4-mL scintillation vials (Skatron, Lier, Norway) using a Titertek disc transfer system (Flow) and 2 mL ofOptiphase X (Fisons, Loughborough, United Kingdom) added to each vial. Radioactivity was counted in a beta counter. The results are expressed as a stimulation index (SI) of the quintuplicate cultures: SI
=
Median counts per minute in mixed cultures Median counts per minute in wife's cells cultures RESULTS Ovulation Induction The response to stimulation is shown in Table 1. In the CC cycles, the women had up to two follicles ::2:: 17 mm diameter, and in the hMG cycles there were up to four follicles of that diameter. Insemination Four women were inseminated with 10 X 106 (8%).
Fertility and Sterility
Table 1 Response to Ovulation Induction in Direct IP Insemination and Control Cycles No. of follicles > 17 mm 1 2 3 4 All
cc
HMG
Treatment
Control
7 (87.5)" 1 (12.5)
6 (66.7) 3 (33.3)
8 (100)
9 (100)
Treatment 6 5 3 3 17
(35.3) (29.4) (17.6) (17.6) (100)
Control 5 4 4 3 16
(31.2) (25.0) (25.0) (18.7) (100)
second cycles because of refusal after direct IP insemination and pregnancy in a control cycle. Odds ratio analysis of the outcome gave an odds ratio for the first cycles of 0.226 (95% confidence interval [CI] 0.001 to 3.435) and for the second cycles of 0.193 (95% CI 0.016 to 2.384), indicating no statistically significant difference between the direct IP insemination and control cycles in pregnancy outcome. Responses of MLR in Fertile Control Patients
• Values in parentheses are percents.
Sperm Recovery
Eleven women agreed to sampling of the PF 24 hours after direct IP insemination. It was possible to identify spermatozoa in all cases, but motile spermatozoa were found in only 3 women. One of these was a 33-year-old woman with unexplained infertility of 6 years' duration who conceived in the subsequent control cycle. Paired Cycles
Seventeen women had their first cycle as a treatment (direct IP insemination) cycle, and in 9 cases it was a control cycle. Of the 26 women who had a first cycle, 24 participated in a second cycle in which the opposite was used. The 2 who did not have second cycles were a woman who found her first cycle (hMG-stimulated treatment cycle) to be stressful and a woman who conceived in her first cycle (CCstimulated control cycle). Pregnancy
In the first cycles, the 17 direct IP insemination cycles (10 hMG; 7 CC) yielded no pregnancies, but the 9 control cycles (7 hMG; 2 CC) gave one in a CC cycle. In the second cycles, the 8 direct IP insemination cycles (7 hMG; 1 CC) again yielded no pregnancies, but the 16 control cycles (9 hMG; 7 CC) gave three pregnancies (2 on hMG and 1 on CC). Overall, there were four pregnancies in the 26 couples (15.4%), and all were in the control group. The ages of these women ranged between 33 and 37 years. Three had unexplained infertility, and one had a poor postcoital test. First and second cycles are not entirely comparable because in a second cycle there has been recent ovulation stimulation and two couples did not have
Vol. 57, No. 1, January 1992
The MLR responses (SI) in the six fertile couples in natural cycles are shown in Table 2. There was no significant difference between the SI for the women and their partners and the SI for the women and unrelated controls (Wilcoxon's test for pair differences). There was no significant change in the SI comparing the three phases of the menstrual cycle. Responses of MLR in Direct IP Insemination Patients
The MLR responses (SI) for the direct IP insemination treatment cycles are shown in Table 3. Before insemination being performed, there was no evidence of the women being sensitized to their partner compared with the unrelated controls. After IP insemination, there was no evidence of increased sensitization to partner's cells. DISCUSSION
Infertile couples who consider assisted reproduction are faced with a range of treatment options, ranging from the minimal invasion of lUI of washed spermatozoa through increasingly invasive methods such as direct IP insemination to IVF and G1FT. There is a clear need to validate such new treatments
Table 2 Mixed Lymphocyte Reaction During the Menstrual Cycle in Six Fertile Couples• Stimulation index
wxu• Menstrual phase Late follicular phase Midluteal phase
22.0 (15.3 to 24.6)
20.7 (14.3 to 31.7)
17.8 (8.2 to 29.6) 21.0 (7.4 to 50.0)
18.5 (10.5 to 25.4) 23.8 (14.1 to 43.0)
• Mixed lymphocyte reaction is expressed as stimulation index median with range in parentheses. b W X H, wife's cells cultured with partner's cells. • W X U, wife's cells cultured with unrelated control's cells.
Campos-Liete et al.
Direct IP insemination
171
Table 3 Mixed Lymphocyte Reaction Response in HMGstimulated Direct IP Insemination Cycles a Stimulation index
CC cycles preinsemination Day 1 after insemination Day 7 after insemination HMG cycles preinsemination Day 1 after insemination Day 7 after insemination
WXHb
wxuc
16.0 (5.4 to 30.8)
27.9 (6.2 to 46.6)
16.6 (4.6 to 31.7)
10.2 (4. 7 to 56.6)
11.0 (3.8 to 23.3)
21.8 (4.8 to 48.6)
22.3 ( 1.0 to 92. 7)
18.7 (4.3 to 58.9)
14.7 (2.6 to 95.8)
12.7 (4.4 to 93.9)
13.0 (13. 7 to 42.9)
17.5 (7.6 to 60.2)
a Mixed lymphocyte reaction is expressed as stimulation index median with range in parentheses. b W X H, wife's cells cultured with partner's cells. c W X U, wife's cells cultured with unrelated control's cells.
in controlled studies to ensure that the additional intervention can indeed be justified. We chose as a control group the same women having the same ovarian stimulation and timed intercourse so that the only element of difference from the direct IP insemination group was the insemination procedure itself. Breaching the integrity of the female reproductive tract by direct IP insemination raises theoretical concerns that it might stimulate an immune response to the spermatozoal antigenic stimulus. This question was raised and dismissed- by Ahlgren.U Ibrahim et al. 12 have reported that humoral immunity was not stimulated because antisperm antibodies were not detected after direct IP insemination. We have assessed whether cell-mediated immune sensitization might occur with direct IP insemination. Using the MLR, there was no difference in the response over the hormonal transition from the low (menstrual) to the high (late follicular) estrogen state then to the estrogen/progesterone (luteal) state. It was reassuring to find that the direct IP insemination cycles were not associated with an increase in MLR sensitization. Advocates of direct IP insemination have reported success with spermatozoal numbers in the same range as were inseminated in the present study, and we confirmed that the needle tip was indeed in the peritoneal cavity in each case. In those women in whom recovery of PF was attempted after 24 hours, we were able to confirm the presence of spermatozoa. Despite these confirmations, we observed no preg-
172
Campos-Liete et al.
Direct IP insemination
nancies in the 25 direct IP insemination cycles. In contrast, there were four pregnancies in the 25 control cycles, giving a clinical pregnancy rate (PR) of 16% per cycle. This rate is similar to that quoted for direct IP insemination in previous uncontrolled studies (14.3% in 56 cycles,2 12.2% in 171 cycles,3 6% in 51 cycles, 7 and 13.2% in 429 cycles5 ). Having carried out the substantial work of recruiting 26 couples to participate in a controlled study, we were faced with the dilemma that there was little prospect of the study demonstrating that direct IP insemination was an effective intervention when the control group already had four pregnancies and the direct IP insemination group none. We estimated that even if the rest of the direct IP insemination series reproduced the 16% PR quoted by Forrler et al. 2 and the control group had no further pregnancies, it would take an additional 75 paired cycles to demonstrate a statistically significant benefit from using direct IP insemination. We realized that recruitment to the controlled study would also become more difficult because couples would naturally inquire as to the success of direct IP insemination at our center. We decided that the trial should be discontinued, and because we had been unable to demonstrate direct IP insemination to be an effective intervention, it was dropped as a treatment option. It is important to add that our failure to achieve direct IP insemination pregnancies was unlikely to relate to inexperience in the handling of gametes because in our IVF work we have achieved a clinical PR of 26% for 1,000 consecutive IVF treatment cycles. Recently, attention has been drawn to the greater tendency for only positive trial results to achieve publication 13 ; the resulting tendency is for these interventions to be viewed as more effective than a balanced view of positive and negative studies would suggest. We believe it is important that our experience is reported because, as far as we can determine, it is the first randomized controlled assessment of direct IP insemination. This study should encourage those contemplating the use of direct IP insemination to reflect on whether it is a worthwhile procedure and should stimulate advocates of direct IP insemination to validate it in controlled studies.
REFERENCES 1. Forrler A, Dellenbach P, Nisand I, Moreau L, Cranz C, Clavert
A, Rumpler Y: Direct intraperitoneal insemination in unexplained and cervical infertility. Lancet 2:916, 1986
Fertility and Sterility
2. Forrler A, Badoc E, Moreau L, Dellenbach P, Cranz C, Clavert A, Rumpler Y: Direct intraperitoneal insemination: first results confirmed. Lancet 2:1468, 1986 3. Menard A, Moreau L, Arbogast E, Dellenbach P: Intraperitoneal insemination: a new method of treatment of various types of sterility. Rev Fr Gynecol Obstet 83:625, 1988 4. Curson R, Parsons J: Disappointing results with direct intraperitoneal insemination. Lancet 1:112, 1987 5. Cimino C, Guastella G, Comparetto G, Gullo D, Perino A, Benigno MA, Barba G, Cittadini E: Direct intraperitoneal insemination (DIP I) for the treatment of refractory infertility unrelated to female organic disease. Acta Eur Fertil 19:61, 1988 6. Sher G, Jacobs MH, Vaught LK, Vaught WG: Intrauterine insemination (lUI) versus lUI with transperitoneal insemination (TPI) in patients with unexplained infertility: comparing clomiphene-vs human menopausal gonadotropin (HMG)-induced controlled ovarian hyperstimulation (COH). (Abstr.) Presented at the Sixth World Congress of In Vitro Fertilization, Jerusalem, Israel, April2 to 7, 1989. Published by the Organizing Committee, in the Congress Proceedings, 1989, p 49
Vol. 57, No.1, January 1992
7. Dooley M, Lim Howe D, Savvas M, Studd JWW: Early experience with gamete intrafallopian transfer (GIFT) and direct intraperitoneal insemination (DIP!). J R Soc Med 81: 637, 1988 8. Templeton A, Aitken J, Mortimer D, Best F: Sperm function in patients with unexplained infertility. Br J Obstet Gynaecol 89:550, 1982 9. World Health Organization: WHO Laboratory Manual for the Examination of Human Semen and Semen-Cervical Mucus Interaction, 2nd edition. Cambridge, The Press Syndicate of the University of Cambridge, 1987, p 3 10. Sargent I, Redman C: Maternal immune responses to the fetus in human pregnancy. In Immunology of Pregnancy and its Disorders, Edited by C Sten. London: Kluwer, 1988, p 115 11. Ahlgren M: Sperm transport to and survival in the human fallopian tube. Gynecol Invest 6:206, 1975 12. Ibrahim ZIZ, Lowe B, Matson PL, Lieberman BA: Antisperm antibodies in infertility. Br Med J 296:934, 1988 13. Chalmers 1: Underreporting research is scientific misconduct. JAMA 263:1405, 1990
Campos-Liete et al.
Direct IP insemination
173
FERTILITY AND STERILITY
Printed on acid-free paper in U.S.A.
Copyright e 1992 The American Fertility Society
A controlled assessment of direct intraperitoneal insemination
Eduardo Campos-Liete, Ph.D.*t Mark lnsull, M.B., Ch.B. Stephen H. Kennedy, M.B., Ch.B.
J. Douglas Ellis, M.B., Ch.B. Ian Sargent, Ph.D. David H. Barlow, M.D.:j:
The Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospita~ Oxford, United Kingdom
Objective: To test the efficacy of direct intraperitoneal (IP) insemination and its effect on cellmediated immunity. Design: A prospective trial with each couple having one treatment (insemination) cycle and one control (timed intercourse) cycle performed in random order with the same ovulation stimulation in both cycles. Setting: Secondary and tertiary referral fertility clinics; university teaching hospital. Patients: Twenty-six infertile couples. Excluding pregnancy, only one couple did not complete the two cycles. Interventions: Ovulation induction in both cycles. Intraperitoneal insemination in the insemination cycles. Main Outcome Measures: These were pregnancy rate (treatment versus control) and mixed lymphocyte response (MLR) sensitivity (before and after direct IP insemination treatment). Results: There were four control and no treatment pregnancies. This was not a significant difference (odds ratio). Mixed lymphocyte responses in fertile subjects did not change during the menstrual cycle (Wilcoxon). There was no significant increase in MLR sensitivity to partner's cells after direct IP insemination treatment. Conclusions: This controlled study found no benefit from direct IP insemination in terms of pregnancies over control cycles. There was no evidence that direct IP insemination had increased Fertil Steril1992;57:168-73 cell-mediated immune response sensitivity to husband's cells.
Direct intraperitoneal (IP) insemination has been promoted as a form of assisted reproduction involving less technology than full-scale in vitro fertilization (IVF) or gamete intrafallopian transfer (GIFT). It was first reported by Forrler et al. 1 and has been applied particularly in cases of unexplained or cervical infertility. 2- 5 It has also been combined with intrauterine insemination (IUI).6-7
Received November 1, 1990; revised and accepted August 23, 1991. * Recipient of a Post-Graduate Education Federal Agency (CAPES) scholarship no. 6859, Ministry of Education, Brasilia, Brazil. t Supported by the Oxford Regional Health Authority Locally Organised Research Scheme, Oxford. *Reprint requests: David H. Barlow, M.D., Nuffield Department of Obstetrics and Gynaecology, Level 3 John Radcliffe Hospital, Headington, Oxford OX3 9DU, United Kingdom.
168
Campos-Liete et al.
Direct IP insemination
The procedure involves injecting motile spermatozoa, after washing and migration, into the pouch of Douglas (cul-de-sac) via the vagina in a woman who has undergone ovulation stimulation. Ovulation is induced by clomiphene citrate (CC) or gonadotropins, and the insemination is timed to coincide with ovulation. The method avoids exposure of the spermatozoa to the cervical mucus and the uterine secretions as well as facilitating their transportation into the peritoneal cavity, a site they are known to reach after sexual intercourse. 8 In bypassing the usual route of entry into the pelvis, it may be that these spermatozoa represent a greater antigentic stimulus that could lead to activation of female cellmediated immune defenses. When new infertility treatments are introduced, it is important that they should undergo validation in controlled trials. Where the treatment involves
Fertility and Sterility
subjecting the woman to ovulation induction and an invasive pelvic procedure, it is particularly important to demonstrate that the treatment is effective. This study assesses the efficacy of direct IP insemination in a controlled trial with each couple having a direct IP insemination cycle and a control cycle and examines the effect of direct IP insemination on the female cell-mediated immune sensitivity to partner's cells. MATERIALS AND METHODS Patients
Twenty-six couples attending the two infertility clinics at the John Radcliffe Hospital, Oxford, United Kingdom, participated in the trial. All had undergone full investigation of infertility, and all had normal findings on laparoscopy within the previous 2 years. Fourteen couples had unexplained infertility, 8 had cervical factor infertility, and 4 had male subfertility. Fifteen had primary infertility (57.6%) and 11 secondary infertility (42.3%). The mean age of the women was 34.3 (range 26 to 40) years. The duration of infertility ranged from 2 to 10 years, with 5 of the 26 couples (19.2%) trying to conceive for >5 years. Study Design
The study design was for each couple to have a treatment and a control cycle in random order, unless pregnancy occurred in the first cycle. A treatment cycle involved ovulation induction with monitoring and timed direct IP insemination. A control cycle involved the same ovulation induction with monitoring and timed intercourse. The ovulation induction was either CC (Serophene; Serono Laboratories, Welwyn Garden City, United Kingdom) 50 to 150 mg/d for 5 days from cycle day 2 or human menopausal gonadotropin (hMG, Pergonal; Serono Laboratories) two to seven ampules on alternate days from cycle day 7. The final duration of administration of hMG was determined by ultrasonic monitoring. Allocation of the women to the two forms of ovulation induction was determined solely by the clinic of origin of the woman. Seventeen women were from the clinic of J.D.E. and received hMG, and 9 women were from the clinic of D.H.B. and received CC. The monitoring and timing were managed by a single researcher (E.C.-L.) for all patients.
Vol. 57, No. 1, January 1992
Human chorionic gonadotropin (hCG, 5,000 IU) was administered when at least one follicle was ;:::17 mm diameter. Only when the decision to administer hCG had been reached was randomization to a treatment or control cycle determined by drawing a card. The couples were asked to refrain from sexual intercourse from the start of ultrasonic monitoring. In control cycles, couples were advised to have sexual intercourse 24 and 48 hours after hCG. In treatment cycles, direct IP insemination was performed 36 to 48 hours after hCG. In the treatment cycles, the peritoneal fluid (PF) environment was sampled 24 hours after direct IP insemination to attempt detection of the inseminated sample. The study design had been passed by the Central Oxford Research Ethics Committee. Semen Preparation
Semen samples were obtained by masturbation and were allowed to liquefy (up to 30 minutes). In 4 cases (15.4%), the sperm density was in the range 10 to 40 X 106/mL; in 14 (53.8%), it was 41 to 100 X 106/mL, and in 8 (30.8%) >100 X 106/mL. A minimum of 5% and a maximum of 65% progressive motility was observed. In 2 cases, the morphology was poor with only 15% aild 19% normal forms. In the other samples, there was 51% to 69% normal morphology by World Health Organization9 criteria. After counting, the semen sample was split and each washed twice in Ham's F-10 medium (GIBCO, Paisley, United Kingdom) including 10% heat-inactivated patient's serum. The pellet containing the sperm was overlayered with medium and incubated at 37°C for 30 minutes. The best of the swim-up samples was used for the insemination. The semen preparations were all carried out by a single researcher (E.C.-L.). Insemination
Under vaginal ultrasound guidance an 18-gauge needle was inserted into the pouch of Douglas (culde-sac) via the vaginal probe needle guide. The needle was targeted on a pool of PF and collection of a portion of the fluid attempted (maximum 5 mL). Where PF could not be aspirated, fluid was allowed to drip from the needle end, confirming that the needle tip was indeed correctly located in the peritoneal cavity. The spermatozoa were injected in 1 mL of medium under ultrasonic visualization. The patient was allowed to rest for at least 15 minutes after the insemination.
Campos-Liete et al.
Direct IP insemination
169
Sperm Recovery Procedure The needling procedure was repeated at 24 hours after the insemination to attempt retrieval of some inseminated sperm. Because we expected that the women would be reluctant to have extra nontherapeutic needling procedures, we limited these to direct IP insemination cycles only. Eleven women agreed to this additional procedure in their direct IP insemination cycles. The volume of recovered fluid was measured and the specimen separated into cellular and liquid components by centrifugation. After washing in phosphate-buffered saline (PBS), a few drops of the cellular suspension were examined under phase-contrast microscopy for the presence of motile spermatozoa and the rest of the suspension further washed in distilled water and spotted (3 spots of 6 J.LL) onto slides. The slides were air dried and fixed using absolute alcohol, mounted with glycerine jelly, and examined under phase-contrast microscopy to determine the number of spermatozoa in each drop. Mixed Lymphocyte Reaction Testing The female cell-mediated immune response was assessed using the mixed lymphocyte reaction (MLR) in which lymphocytes from the woman (recipient) and her partner or unrelated male (donor) are cultured together for 6 days. The donor (stimulator) cells are irradiated to prevent them from responding. The proliferation of recipient (responder) cells is assessed via the incorporation of tritiated thymidine into newly synthesized deoxyribonucleic acid, and in a graft response this would peak at 6 days. By comparing the MLR reactions between the woman and her partner and between the woman and an unrelated control enhancement or suppression of the woman's immune reactivity may be tested. 10 Mixed lymphocyte reaction responses were assessed before direct IP insemination, 1 day after insemination and 7 days after insemination. Couples undergoing IP insemination were assessed in CCstimulated cycles (n = 8) or hMG-stimulated cycles (n = 17). A control group of fertile couples were assessed during natural menstrual cycles (n = 6) to determine if the major hormonal transitions of the menstrual cycle affected MLR responses. Each of the control couples had proven fertility (a child in the past 3 years), and the women had regular ovulatory menstrual cycles of 26 to 32 days' duration. They were not using hormonal contraception. Both partners gave blood samples on the 1st
170
Campos-Liete et al.
Direct IP insemination
day of the menstrual cycle, in the late follicular phase, and in the midluteal phase. Technique of MLR The peripheral blood samples were diluted 1:1 with PBS (10 mL) containing 5 J.Lg/mL heparin. The mononuclear cells were separated on Ficoll-Paque (Pharmacia, Upsalla, Sweden) and resuspended at 1 X 106 cells/mL in RPMI 1640 medium (Flow, Irvine, United Kingdom) with 25 mM Hepes and Lglutamine (GIBCO), containing antibiotic (50 J.Lgf mL penicillin and 50 J.Lg/mL streptomycin) and 10% pooled heat-inactivated AB serum from healthy donors. Quintuplicate 200-J.LL cultures were set up in flatbottomed 96-well microtiter plates with cellular mixtures of responder and stimulator cells, ratio 1: 1. The stimulator cells from husband and unrelated control were inactivated by irradiation (2,000 rad). The cell cultures were incubated at 37°C and 5% C0 2 in air. Blastogenesis was evaluated 5 days later, by adding 0.5 mCi [3H]thymidine to the wells. After a further 24-hour incubation, the cultures were harvested onto glass fiber filters that were allowed to dry. The filters were transferred to 4-mL scintillation vials (Skatron, Lier, Norway) using a Titertek disc transfer system (Flow) and 2 mL ofOptiphase X (Fisons, Loughborough, United Kingdom) added to each vial. Radioactivity was counted in a beta counter. The results are expressed as a stimulation index (SI) of the quintuplicate cultures: SI
=
Median counts per minute in mixed cultures Median counts per minute in wife's cells cultures RESULTS Ovulation Induction The response to stimulation is shown in Table 1. In the CC cycles, the women had up to two follicles ::2:: 17 mm diameter, and in the hMG cycles there were up to four follicles of that diameter. Insemination Four women were inseminated with 10 X 106 (8%).
Fertility and Sterility
Table 1 Response to Ovulation Induction in Direct IP Insemination and Control Cycles No. of follicles > 17 mm 1 2 3 4 All
cc
HMG
Treatment
Control
7 (87.5)" 1 (12.5)
6 (66.7) 3 (33.3)
8 (100)
9 (100)
Treatment 6 5 3 3 17
(35.3) (29.4) (17.6) (17.6) (100)
Control 5 4 4 3 16
(31.2) (25.0) (25.0) (18.7) (100)
second cycles because of refusal after direct IP insemination and pregnancy in a control cycle. Odds ratio analysis of the outcome gave an odds ratio for the first cycles of 0.226 (95% confidence interval [CI] 0.001 to 3.435) and for the second cycles of 0.193 (95% CI 0.016 to 2.384), indicating no statistically significant difference between the direct IP insemination and control cycles in pregnancy outcome. Responses of MLR in Fertile Control Patients
• Values in parentheses are percents.
Sperm Recovery
Eleven women agreed to sampling of the PF 24 hours after direct IP insemination. It was possible to identify spermatozoa in all cases, but motile spermatozoa were found in only 3 women. One of these was a 33-year-old woman with unexplained infertility of 6 years' duration who conceived in the subsequent control cycle. Paired Cycles
Seventeen women had their first cycle as a treatment (direct IP insemination) cycle, and in 9 cases it was a control cycle. Of the 26 women who had a first cycle, 24 participated in a second cycle in which the opposite was used. The 2 who did not have second cycles were a woman who found her first cycle (hMG-stimulated treatment cycle) to be stressful and a woman who conceived in her first cycle (CCstimulated control cycle). Pregnancy
In the first cycles, the 17 direct IP insemination cycles (10 hMG; 7 CC) yielded no pregnancies, but the 9 control cycles (7 hMG; 2 CC) gave one in a CC cycle. In the second cycles, the 8 direct IP insemination cycles (7 hMG; 1 CC) again yielded no pregnancies, but the 16 control cycles (9 hMG; 7 CC) gave three pregnancies (2 on hMG and 1 on CC). Overall, there were four pregnancies in the 26 couples (15.4%), and all were in the control group. The ages of these women ranged between 33 and 37 years. Three had unexplained infertility, and one had a poor postcoital test. First and second cycles are not entirely comparable because in a second cycle there has been recent ovulation stimulation and two couples did not have
Vol. 57, No. 1, January 1992
The MLR responses (SI) in the six fertile couples in natural cycles are shown in Table 2. There was no significant difference between the SI for the women and their partners and the SI for the women and unrelated controls (Wilcoxon's test for pair differences). There was no significant change in the SI comparing the three phases of the menstrual cycle. Responses of MLR in Direct IP Insemination Patients
The MLR responses (SI) for the direct IP insemination treatment cycles are shown in Table 3. Before insemination being performed, there was no evidence of the women being sensitized to their partner compared with the unrelated controls. After IP insemination, there was no evidence of increased sensitization to partner's cells. DISCUSSION
Infertile couples who consider assisted reproduction are faced with a range of treatment options, ranging from the minimal invasion of lUI of washed spermatozoa through increasingly invasive methods such as direct IP insemination to IVF and G1FT. There is a clear need to validate such new treatments
Table 2 Mixed Lymphocyte Reaction During the Menstrual Cycle in Six Fertile Couples• Stimulation index
wxu• Menstrual phase Late follicular phase Midluteal phase
22.0 (15.3 to 24.6)
20.7 (14.3 to 31.7)
17.8 (8.2 to 29.6) 21.0 (7.4 to 50.0)
18.5 (10.5 to 25.4) 23.8 (14.1 to 43.0)
• Mixed lymphocyte reaction is expressed as stimulation index median with range in parentheses. b W X H, wife's cells cultured with partner's cells. • W X U, wife's cells cultured with unrelated control's cells.
Campos-Liete et al.
Direct IP insemination
171
Table 3 Mixed Lymphocyte Reaction Response in HMGstimulated Direct IP Insemination Cycles a Stimulation index
CC cycles preinsemination Day 1 after insemination Day 7 after insemination HMG cycles preinsemination Day 1 after insemination Day 7 after insemination
WXHb
wxuc
16.0 (5.4 to 30.8)
27.9 (6.2 to 46.6)
16.6 (4.6 to 31.7)
10.2 (4. 7 to 56.6)
11.0 (3.8 to 23.3)
21.8 (4.8 to 48.6)
22.3 ( 1.0 to 92. 7)
18.7 (4.3 to 58.9)
14.7 (2.6 to 95.8)
12.7 (4.4 to 93.9)
13.0 (13. 7 to 42.9)
17.5 (7.6 to 60.2)
a Mixed lymphocyte reaction is expressed as stimulation index median with range in parentheses. b W X H, wife's cells cultured with partner's cells. c W X U, wife's cells cultured with unrelated control's cells.
in controlled studies to ensure that the additional intervention can indeed be justified. We chose as a control group the same women having the same ovarian stimulation and timed intercourse so that the only element of difference from the direct IP insemination group was the insemination procedure itself. Breaching the integrity of the female reproductive tract by direct IP insemination raises theoretical concerns that it might stimulate an immune response to the spermatozoal antigenic stimulus. This question was raised and dismissed- by Ahlgren.U Ibrahim et al. 12 have reported that humoral immunity was not stimulated because antisperm antibodies were not detected after direct IP insemination. We have assessed whether cell-mediated immune sensitization might occur with direct IP insemination. Using the MLR, there was no difference in the response over the hormonal transition from the low (menstrual) to the high (late follicular) estrogen state then to the estrogen/progesterone (luteal) state. It was reassuring to find that the direct IP insemination cycles were not associated with an increase in MLR sensitization. Advocates of direct IP insemination have reported success with spermatozoal numbers in the same range as were inseminated in the present study, and we confirmed that the needle tip was indeed in the peritoneal cavity in each case. In those women in whom recovery of PF was attempted after 24 hours, we were able to confirm the presence of spermatozoa. Despite these confirmations, we observed no preg-
172
Campos-Liete et al.
Direct IP insemination
nancies in the 25 direct IP insemination cycles. In contrast, there were four pregnancies in the 25 control cycles, giving a clinical pregnancy rate (PR) of 16% per cycle. This rate is similar to that quoted for direct IP insemination in previous uncontrolled studies (14.3% in 56 cycles,2 12.2% in 171 cycles,3 6% in 51 cycles, 7 and 13.2% in 429 cycles5 ). Having carried out the substantial work of recruiting 26 couples to participate in a controlled study, we were faced with the dilemma that there was little prospect of the study demonstrating that direct IP insemination was an effective intervention when the control group already had four pregnancies and the direct IP insemination group none. We estimated that even if the rest of the direct IP insemination series reproduced the 16% PR quoted by Forrler et al. 2 and the control group had no further pregnancies, it would take an additional 75 paired cycles to demonstrate a statistically significant benefit from using direct IP insemination. We realized that recruitment to the controlled study would also become more difficult because couples would naturally inquire as to the success of direct IP insemination at our center. We decided that the trial should be discontinued, and because we had been unable to demonstrate direct IP insemination to be an effective intervention, it was dropped as a treatment option. It is important to add that our failure to achieve direct IP insemination pregnancies was unlikely to relate to inexperience in the handling of gametes because in our IVF work we have achieved a clinical PR of 26% for 1,000 consecutive IVF treatment cycles. Recently, attention has been drawn to the greater tendency for only positive trial results to achieve publication 13 ; the resulting tendency is for these interventions to be viewed as more effective than a balanced view of positive and negative studies would suggest. We believe it is important that our experience is reported because, as far as we can determine, it is the first randomized controlled assessment of direct IP insemination. This study should encourage those contemplating the use of direct IP insemination to reflect on whether it is a worthwhile procedure and should stimulate advocates of direct IP insemination to validate it in controlled studies.
REFERENCES 1. Forrler A, Dellenbach P, Nisand I, Moreau L, Cranz C, Clavert
A, Rumpler Y: Direct intraperitoneal insemination in unexplained and cervical infertility. Lancet 2:916, 1986
Fertility and Sterility
2. Forrler A, Badoc E, Moreau L, Dellenbach P, Cranz C, Clavert A, Rumpler Y: Direct intraperitoneal insemination: first results confirmed. Lancet 2:1468, 1986 3. Menard A, Moreau L, Arbogast E, Dellenbach P: Intraperitoneal insemination: a new method of treatment of various types of sterility. Rev Fr Gynecol Obstet 83:625, 1988 4. Curson R, Parsons J: Disappointing results with direct intraperitoneal insemination. Lancet 1:112, 1987 5. Cimino C, Guastella G, Comparetto G, Gullo D, Perino A, Benigno MA, Barba G, Cittadini E: Direct intraperitoneal insemination (DIP I) for the treatment of refractory infertility unrelated to female organic disease. Acta Eur Fertil 19:61, 1988 6. Sher G, Jacobs MH, Vaught LK, Vaught WG: Intrauterine insemination (lUI) versus lUI with transperitoneal insemination (TPI) in patients with unexplained infertility: comparing clomiphene-vs human menopausal gonadotropin (HMG)-induced controlled ovarian hyperstimulation (COH). (Abstr.) Presented at the Sixth World Congress of In Vitro Fertilization, Jerusalem, Israel, April2 to 7, 1989. Published by the Organizing Committee, in the Congress Proceedings, 1989, p 49
Vol. 57, No.1, January 1992
7. Dooley M, Lim Howe D, Savvas M, Studd JWW: Early experience with gamete intrafallopian transfer (GIFT) and direct intraperitoneal insemination (DIP!). J R Soc Med 81: 637, 1988 8. Templeton A, Aitken J, Mortimer D, Best F: Sperm function in patients with unexplained infertility. Br J Obstet Gynaecol 89:550, 1982 9. World Health Organization: WHO Laboratory Manual for the Examination of Human Semen and Semen-Cervical Mucus Interaction, 2nd edition. Cambridge, The Press Syndicate of the University of Cambridge, 1987, p 3 10. Sargent I, Redman C: Maternal immune responses to the fetus in human pregnancy. In Immunology of Pregnancy and its Disorders, Edited by C Sten. London: Kluwer, 1988, p 115 11. Ahlgren M: Sperm transport to and survival in the human fallopian tube. Gynecol Invest 6:206, 1975 12. Ibrahim ZIZ, Lowe B, Matson PL, Lieberman BA: Antisperm antibodies in infertility. Br Med J 296:934, 1988 13. Chalmers 1: Underreporting research is scientific misconduct. JAMA 263:1405, 1990
Campos-Liete et al.
Direct IP insemination
173
